[ATM] Report from a professional optical shop near Baltimore

[Guy Brandenburg <gfbrandenburg@yahoo.com>]

On our way to pick up a bunch of 8-foot-long, 10-inch
ID, 1/4 inch cardboard sonotubes for telescope tubes,
a friend and I stopped by an optical manufacturing
shop just north of Baltimore, at the invitation of one
of its managers, so we could take a look at their
methods of grinding, polishing, and figuring mirrors
and other optical items. (I was asked not to identify
the company.) They had all sorts of milling machines
that made short work of the rough grinding of the
general shape of the mirrors, some of which were very
large (4 feet across or more) and very exotic (lots of
them were made out of Cer-Vit or Zerodur), and paid
for mostly by defense contractors. The milling
machines had sort of a cup on the working end that had
diamond teeth. This gizmo was set some sort of an
angle to grind out the proper curvature, I surmise,
although we didn't see any of them in operation.

We saw a case where there was a rather steeply-curved
mirror that a worker was measuring and fine-grinding
to some sort of an exotic shape in a most unusual
fashion. The measurement device consisted of a long,
carefully-measured vertical steel holder that was
about 8 to 10 feet in the air over the place where the
measurement took place; there was a heavy, accurate
metal braided cable attached to that high point; and
at the end of that was a dial indicator. The worker
put the mirror into its test bed directly below the
pivot point. Then, using a specially-made wooden
stick, the worker moved the dial indicator over the
surface of the mirror and noted directly on the mirror
with a Sharpie magic marker what the numbers were at
various locations thereon. Then he took the mirror
back to the grinding machine for more fine-grinding,
where he presumably adjusted the machine to cause some
desired effect.

They had very large (4 to 8 foot diameter)  planetary
turntables that polished flats against a rotating lap.
The way this worked was that there were very large
metal cylinders that had walls about an inch thick and
4 or 5 inches tall and about 12 to 18 inches interior
diameter. These cylinders then held the mirror (of
whatever shape it might be) on the turntable, but also
rotated themselves somehow, forcing the mirrors to
polish their faces against the lap at lots of
different orientations. I forget exactly how the
cylinders rotated; I didn't take photos or notes. The
laps looked to be either metal gratings or some sort
of really hard pitch, not sure which.

They had lots of motorized turntables to polish or
fine grind mirrors. The turntables worked with the
mirror face up, in general, and an under-sized lap on
top. I would estimate that the stroke of the overarm
was on the range of 1 or 2 cycles per second. Workers
would use a paintbush to paint on some of the
abrasive. They used a good bit of cerium oxide - no
rouge - and generally seemed to purchase their CeO
already packaged as a premixed slurry. 

They had interferometers to judge which parts of a
mirror were high or low in comparison with a reference
mirror. These were rather complicated contraptions
with beam splitters, lasers, collimators, air tables,
computer printouts, and so on. Way over our budget,
believe me. Several of them carried the name 'Zygo'. I
can't begin to really explain how they worked. We
didn't get a chance to see one in operation,
unfortunately; the fellow who specializes in working
on them was on vacation that week.

One thing I did notice is that when they test mirrors,
they don't bother with making masks, and they don't
care about getting it so that each zone has the exact
same area. They simply write directly on the mirror
surface with a magic marker - again - every inch or
half inch or centimeter or whatever, and take readings
at each of those zones, using a standard knife-edge
test just like we do. Only their apparatus is at once
more complex and more simple than what we use. For a
source of light, they almost always use a pinhole at
the end of a fiber-optic strand. They don't have to
use slits. That's the complex part. The simple part?
They just use some glue or wax to bond a double-edge
razor blade to a metal stick, attached to a base that
slides around on the table!

They even had an area where they melted pitch for
various specifications; this looked like a cross
between chemist's hood area and a forge at a
blacksmith's shop. There were probably more than
20 different named varieties of pitch in gallon- to
quart- sized metal cans with various names painted on
them. No burgundy pitch - it almost all looked like
that black Gugolz stuff, but they also had some
synthetic varieties. Apparently, different mirrors for
different purposes require different varieties of
pitch.

We also saw how they got a rough polish on one of
their mirrors (of about 36 inch diameter): they made a
metal tool to fit its curvature, then glued a pad and
then some nylon loop carpet to the tool! 

I also got a brief introduction to a test known as the
wire test, which might be useful when testing really
fast mirrors (below f/5). There is a simulation and
data reduction program that will show what the wire
test should look like for a given zone of a given
mirror at a given distance from the COC. I forget the
name of the program, however. The wire test is quite
interesting. You don't need a mask. You get the wire
in the right area for a given zone, both laterally and
longitudinally, and then you get a nice loop of shadow
around the center of the mirror in the correct zone
area. 

I am sure that I forgot more than I reported. 

I really need to write them a nice thank-you note.

Guy Brandenburg


Guy  Brandenburg
Washington, DC
My home page:
http://home.earthlink.net/~gfbranden/GFB_Home_Page.html
_______________________________________________
ATM mailing list http://www.atmlist.net/